Isolation of Plakortolides - American Chemical Society

Jan 24, 2011 - Queensland Museum, P.O. Box 3300, South Brisbane, QLD 4101, Australia. bS Supporting Information. ABSTRACT: Sixteen new compounds ...
1 downloads 0 Views 2MB Size
ARTICLE pubs.acs.org/jnp

Configurational Assignment of Cyclic Peroxy Metabolites Provides an Insight into Their Biosynthesis: Isolation of Plakortolides, seco-Plakortolides, and Plakortones from the Australian Marine Sponge Plakinastrella clathrata Ken W. L. Yong,† James J. De Voss,† John N. A. Hooper,‡ and Mary J. Garson*,† † ‡

School of Chemistry and Molecular Biosciences, The University of Queensland, Brisbane, QLD 4072, Australia Queensland Museum, P.O. Box 3300, South Brisbane, QLD 4101, Australia

bS Supporting Information ABSTRACT:

Sixteen new compounds, comprising nine new plakortolides K-S (1-9), four seco-plakortolides (10-13), and three plakortones (14-16), were isolated from the Australian sponge Plakinastrella clathrata. Structural elucidation, including relative configurational assignment, was based on extensive spectroscopic analysis, while the absolute configurations of 1-4 were deduced from 1H NMR analyses on MPA esters derived from Zn/AcOH reduction products. Diastereomeric sets of plakortolides, e.g., K and L, or M and N, differed in configuration at C-3/C-4 rather than at C-6, a stereochemical result that compromises a biosynthetic pathway involving Diels-Alder cycloaddition of molecular oxygen to a Δ3,5-diunsaturated fatty acid precursor. The biosynthesis may plausibly involve cyclization of a 6-hydroperoxydienoic acid intermediate following stereospecific introduction of the hydroperoxy group into a polyketide-derived precursor. Isolated seco-plakortolides converted under mild conditions into plakortones with full retention of configuration, suggesting C-6 hydroxy attack on an R,β-unsaturated lactone intermediate. The NMR data reported for the compound named plakortolide E are inconsistent with the current literature structure and are those of the corresponding secoplakortolide (19). The reported conversion of the metabolite into a plakortone ether on storage is consistent with this structural revision.

metabolite named plakortolide included the first detailed description of the assignment of relative configuration and used ROESY data and molecular modeling for this purpose.4 The various plakortolides isolated since these two initial reports exhibit variation in alkyl chain length (generally C8 or C10, but some C9 examples are found), methylation and/or unsaturation pattern, and terminal group (either phenyl or p-hydroxyphenyl).4-11 In this paper we report six new plakortolides (1-6) possessing a C12 methylene chain, three C10 homologues (7-9), together with seco derivatives 10-13, and the stereochemically related plakortones 14-16 from Plakinastrella clathrata Kirkpatrick,

C

yclic peroxides are a distinctive suite of bioactive metabolites frequently encountered in marine sponges of the genera Plakortis and Plakinastrella. Individual metabolites can be categorized as belonging to the plakortolide, plakinic acid, plakoric acid, plakortone, or plakortide families.1 In addition to pronounced cytotoxic, antifungal, and anti-inflammatory activity, recent screening programs have shown that members of these classes of metabolites possess potent antiparasitic activity and so may be promising drug leads against tropical diseases such as malaria, African sleeping sickness, or leishmaniasis.1,2 The plakortolide structure is characterized by an aromatic unit connected via a methylene chain to a bicyclic peroxy-lactone ring system. A metabolite with a plakortolide skeleton was briefly described in 1980,3 but Davidson’s 1991 report of the cytotoxic Copyright r 2011 American Chemical Society and American Society of Pharmacognosy

Received: September 3, 2010 Published: January 24, 2011 194

dx.doi.org/10.1021/np100620x | J. Nat. Prod. 2011, 74, 194–207

Journal of Natural Products

ARTICLE

Chart 1

Plakortolides and seco-Plakortolides. Plakortolide K (1) was isolated as a colorless oil and had a formula of C26H40O4 inferred from HRESIMS. The 1H NMR and HSQC data (Tables 1 and 2) showed a phenyl ring (δH 7.25 (2H), 7.16 (2H), and 7.15 (1H); δC 142.8 (s), 127.8 (d), 128.2 (d), and 125.1 (d)), two methyl groups (δH 1.36 (s), 1.18 (s); δC 25.7, 24.4), one oxygenated methine (δH 4.46; δC 80.5), two methylenes with a diagnostic AB signal pattern (δH 2.88, 2.55; δC 33.7 and δH 2.25, 1.63; δC 40.0), one other identifiable methylene (δH 2.58; δC 35.6), and a cluster of overlapping methylene signals (δH 1.23-1.29; δC 29.0-29.2), suggesting the presence of a long alkyl chain in the structure. HMBC data were all fully consistent with the plakortolide skeleton.4-11 Important correlations were from H-3 (δH 4.46) to a lactone carbonyl (C-1; δC 174.0), to C-2 (δC 33.7), to an oxygenated quaternary carbon (C-4; δC 82.3), and to Me-24 (δC 25.7); from the Me-23 signal at δH 1.36 to C-3 (δC 80.5) and C-5 (δC 40.0); from H-5 (δH 2.25 and 1.63) to Me-23 and Me-24; and from Me-24 (δH 1.18) to C-5 and to C-7 (δC 36.6). HMBC correlations from H-18 (δH 2.58) to an ipso aromatic carbon (C-19; δC 142.8), to C-20 (δC 128.2), and to some individual carbons of the long alkyl chain confirmed that the phenyl ring was attached to the plakortolide skeleton via an alkyl chain, the length of which was C12 by HRESIMS.

1900. Our study addresses the relative and absolute configurations of the suite of metabolites and reveals that in this particular sponge diastereomeric plakortolides such as 1 and 2 differ in configuration at the linked C-3/C-4 centers, rather than at C-6. These stereochemical results are considered in light of plausible biosynthetic pathways leading to the plakortolides. By comparison of our data with those of plakortolide E (17)6 and its diastereomer (18),7,12 we determine that the NMR data previously ascribed to plakortolide E (17) are in fact those of a seco analogue, 19. We also find that seco-plakortolides convert easily into ether products possessing a plakortone skeleton.

’ RESULTS AND DISCUSSION Structural and Stereochemical Studies. A single large specimen of P. clathrata Kirkpatrick, 1900 (Homoscleromorpha) was collected by scuba from the Gneerings Reef offshore from Mooloolaba in South East Queensland. The sponge was extracted with CH2Cl2/MeOH (1:1) to give an extract, which was fractionated by NP flash chromatography (CH2Cl2/EtOAc) followed by NP-HPLC (hexanes/EtOAc) or by RP-HPLC (CH3CN/H2O) to give nine new plakortolides (1-9), the related seco compounds (10-13), and plakortone ethers (14-16). 195

dx.doi.org/10.1021/np100620x |J. Nat. Prod. 2011, 74, 194–207

a

1.23-1.29, m

9

196

1.24-1.29, mc,d

1.23-1.29, m 1.23-1.29, mc,d

1.23-1.29, mc,d

1.59, m

2.58, t (7.8)

7.16, m

7.25, m

7.15, m

1.36, s 1.18, s

14 15

16

17

18

20

21

22

23 24

22-OH

c,d

4.74, br s

1.36, s 1.18, s

6.72, d (8.5)

7.02, d (8.5)

2.50, t (7.8)

1.54, m

1.22-1.27, mc,d

1.22-1.27, m 1.22-1.27, mc,d

c,d

1.22-1.27, m

c,d

1.22-1.27, m

c,d

1.22-1.27, m

c,d

1.22-1.27, m

c,d

1.22-1.27, m

1.25, m

1.71, m 1.54, m

1.64, d (15.0)

2.26, d (15.0)

4.46, d (6.0)

2.55, br d (18.6)

2.89, dd (18.6, 6.0)

3

c,d

4.75, br s

1.37, s 1.26, s

6.72, d (8.5)

7.01, d (8.5)

2.50, t (7.8)

1.54, m

1.23-1.27, mc,d

1.23-1.27, m 1.23-1.27, mc,d

c,d

1.23-1.27, m

c,d

1.23-1.27, m

c,d

1.23-1.27, m

c,de

1.23-1.27, m

c,d

1.23-1.27, m

1.25, m

1.50-1.46, m

1.69, d (15.0)

2.15, d (15.0)

4.43, d (6.0)

2.50, br d (18.6)

2.90, dd (18.6, 6.0)

4

1.35, s 1.18, s

7.16, t (7.5)

7.27, t (8.0)

7.35, d (7.5)

6.42, d (15.5)

6.73, dd (15.5, 10.5)

6.18, dd (15.0, 10.5)

2.12, q (7.0) 5.81, dt (15.0, 7.0)

1.23-1.26, m

c,d

1.23-1.26, m

c,d

1.23-1.26, m

c,d

1.23-1.26, m

c,d

1.23-1.26, m

1.28, m

1.23-1.26, mc,d

1.37, s 1.27, s

7.17, t (7.3)

7.28, t (7.7)

7.35, d (7.3)

6.42, d (15.7)

6.73, dd (15.7, 10.5)

6.18, dd (15.0, 10.5)

2.12, q (7.0) 5.80, dt (15.0, 7.0)

1.25-1.25, m

c,d

1.25-1.25, m

c,d

1.25-1.25, m

c,d

1.25-1.25, m

c,de

1.25-1.25, m

c,d

1.48, m

-c,d

c,d

1.69, d (14.8)

2.15, d (14.8)

4.43, d (6.2)

2.60, br d (18.6)

2.89, dd (18.6, 6.2)

6

1.63, d (15.0)

2.25, d (15.0)

4.46, d (6.0)

2.54, br d (18.5)

2.88, dd (18.5, 6.0)

5

c,d

1.17, s

1.36, s

7.17, t (7.3)

7.22, t (7.7)

7.35, d (7.3)

6.43, d (15.5)

6.18, dd (15.0, 10.5) 6.74, dd (15.5, 10.5)

5.81, dt (15.0, 7.0)

2.11, q, 7.0

1.23-1.25, m

c,d

1.23-1.25, m

c,d

1.23-1.25, m

1.29, m

1.53, m

1.64, d (15.0)

2.26, d (15.0)

4.46, d (6.0)

2.54, br d (18.5)

2.88, dd (18.5, 6.0)

7

c,d

1.27, s

1.37, s

7.17, t (7.3)

7.29, t (7.7)

7.35, d (7.2)

6.42, d (15.7)

6.18, dd (15.1, 10.5) 6.73, dd (15.7, 10.5)

5.79, dt (15.1, 7.2)

2.12, q, 7.2

1.26-1.31, m

c,d

1.26-1.31, m

c,d

1.26-1.31, m

1.30, m

1.52-1.48, m

1.69, d (14.8)

2.15, d (14.8)

4.43, d (6.2)

2.60, br d (18.6)

2.89, dd (18.6, 6.2)

8

9

1.18, s

1.36, s

6.72, d (8.5)

7.01, d (8.5)

2.51, t (7.5)

1.20-1.30, mc,d 1.20-1.30, mc,d

1.20-1.30, mc,d

1.20-1.30, mc,d

1.20-1.30, mc,d

1.20-1.30, mc,d

1.20-1.30, mc,d

1.20-1.30, mc,d

1.53, m

1.64, d (15.0)

2.25, d (15.0)

4.48, d (6.0)

2.55, br d (18.8)

2.89, dd (18.8, 6.0)

Chemical shifts (ppm) referenced to CHCl3 (δC 7.24). b At 500 MHz. c Unresolved chemical shifts due to overlapping signals. d Signal multiplicity unresolved due to overlapping signals.

1.37, s 1.27, s

7.15, m

7.25, m

7.16, m

2.57, t (7.6)

1.59, m

1.24-1.29, m 1.24-1.29, mc,d

1.24-1.29, m c,d

1.23-1.29, m

c,d

13

c,d

c,d

1.24-1.29, m

1.24-1.29, m c,d

1.23-1.29, m

1.23-1.29, m

1.24-1.29, m

c,d

12

11

1.23-1.29, m

1.24-1.29, m

1.27, m

c,d

1.25, m

8

1.50-1.46, m

1.68, d (15.0)

2.15, d (15.0)

c,d

1.71, m 1.54, m

7

10

1.63, d (15.0)

5b

c,d

2.25, d (15.0)

5a

4.43, d (6.1)

c,d

4.46, d (5.9)

3

2.60, br d (18.6)

2.89, dd (18.6, 6.1)

c,d

2.55, br d (18.5)

2

c,d

2.88, dd (18.5, 5.9)

2b

1

2a

position

Table 1. 1H NMR Assignments for Plakortolides 1-9a,b

Journal of Natural Products ARTICLE

dx.doi.org/10.1021/np100620x |J. Nat. Prod. 2011, 74, 194–207

Journal of Natural Products Table 2.

13

ARTICLE

C NMR Assignments for Plakortolides 1-4 and 6-8a,b

position

1

2

3

4

6

7

8

1

174.0

174.1

174.1

174.1

174.1

174.1

174.1

2

33.7

34.2

33.9

34.0

34.2

33.9

34.1

3

80.5

81.0

80.7

80.8

81.0

80.6

80.9

4

82.3

82.5

82.4

82.6

82.6

82.3

82.5

5

40.0

40.4

40.1

40.3

40.5

40.1

40.3

6

80.0

79.9

80.0

80.0

79.9

80.0

79.9

7

36.6

40.8

36.8

40.8

40.7

36.6

40.7

8 9

23.3 29.0-29.2c

22.8 29.3-29.4c

23.5 29.1-29.4c

22.6 29.0-29.2c

22.8 29.0-29.4c

23.4 29.0-29.5c

22.7 28.8-29.6c

10

29.0-29.2c

29.3-29.4c

29.1-29.4c

29.0-29.2c

29.0-29.4c

29.0-29.5c

28.8-29.6c

11

c

c

c

c

c

c

28.8-29.6c

29.0-29.2

c

29.3-29.4

c

29.1-29.4

29.0-29.2

c

c

29.0-29.4

c

29.0-29.5

12

29.0-29.2

29.3-29.4

29.1-29.4

29.0-29.2

29.0-29.4

32.6

32.5

13

29.0-29.2c

29.3-29.4c

29.1-29.4c

29.0-29.2c

29.0-29.4c

135.8

135.5

14

29.0-29.2c

29.3-29.4c

29.1-29.4c

29.0-29.2c

32.7

130.3

130.4

15

29.0-29.2c

29.3-29.4c

29.1-29.4c

29.0-29.2c

135.8

129.4

129.2

16 17

29.0-29.2c 31.3

29.3-29.4c 31.3

29.1-29.4c 31.4

29.0-29.2c 31.5

130.4 129.3

129.7 137.6

129.8 137.5

18

35.6

35.9

34.8

34.6

129.8

125.9

125.9

19

142.8

142.7

135.0

135.0

137.4

128.3

128.3

20

128.2

128.3

129.3

129.2

125.9

126.8

126.9

21

127.8

128.0

114.9

114.7

128.3

25.6

25.7

22

125.1

125.3

153.5

153.3

126.9

24.6

22.2

23

25.7

25.8

25.8

25.6

25.8

24

24.4

22.2

24.6

22.0

22.2

a

Chemical shifts (ppm) taken from 2D spectra referenced to CDCl3 (δC 77.0). b At 500 MHz. c Unresolved chemical shifts taken from HSQC experiments.

Figure 1. Key NOESY correlations for the bicyclic cores of plakortolides K (1) and L (2).

Plakortolide L (2) eluted earlier than 1 from NP-HPLC, but had the same molecular formula based on HRESIMS measurement. HMBC and HSQC data revealed that compound 2 had a plakortolide skeleton, but a comparison of chemical shifts for the bicyclic core revealed subtle differences in the chemical shifts of C-7 and of the C-24 methyl group (C-7, δC 36.6 for 1 vs δC 40.8 for 2; C-24, δH 1.18; δC 24.4 for 1 vs δH 1.27; δC 22.2 for 2) (Tables 1 and 2) that suggested that plakortolides K and L differed in configuration at C-6. Alternatively, they differed in configuration at the C-3/C-4 positions, as had been reported earlier for plakortolide E and its diastereomer by Mosher analysis.6,7 In NOESY experiments (Figure 1), both compounds

1 and 2 showed a strong dipole coupling between H-3 and H3-23, confirming the cis junction between the lactone and peroxide rings that is required on stereochemical grounds, and also between H-3 and H-5b. In 1, both H3-23 and H3-24 showed a strong NOE correlation to H-5b (δH 1.63), while H-5a at δH 2.25 showed an NOE to H-7b and to other protons in the side chain. In his stereochemical analysis of plakortolide, Davidson4 noted that the downfield H-5a signal showed an NOE to the side chain proton H-7b, thereby placing the side chain on the opposite face of both methyls. The C-24 chemical shift data of 1 matched values observed in plakortolide (C-24 δH 1.22; δC 24.93).4 In contrast in 2, H3-23 and H3-24 showed strong NOE 197

dx.doi.org/10.1021/np100620x |J. Nat. Prod. 2011, 74, 194–207

Journal of Natural Products

ARTICLE

Figure 2. Mosher ester analysis of plakortolides K (1) and L (2).

correlations to H-5a at δH 1.68 and to H-5a at δH 2.15, respectively, as does the recently reported plakortolide J. The C-7/C-24 chemical shift data of 2 matched values observed in plakortolide J (C-7 δC 41.0; C-24 δH 1.28; δC 22.4).11 Thus 1 has the same relative configuration as the original plakortolide, and 2 was either a C-6 epimer of 1 or the C-3/C-4 diastereomer. The C-3 configurations of 1 and 2 were determined by reductive cleavage of the peroxy ring using Zn/AcOH to afford samples of diols 10 and 11, which were esterified at C-3 to their (R)- and (S)-O-methyl mandelate (MPA) esters 20a/20b and 21a/21b, respectively.2a,2f,6,7 For the MPA esters 20a/20b from diol 10, the ΔδRS values (where ΔδRS = δR - δS) were positive for H2-2 and negative for H-3, H2-5, H3-23, and H3-24, consistent with a 3R configuration (Figure 2); in contrast, for the MPA esters 21a/21b from diol 11, the ΔδRS values were negative for H2-2 and positive for H2-5, H2-7, H3-23, and H3-24, consistent with a 3S configuration. Together with the relative configurational information, these results established 1 as (3R, 4R, 6S) and 2 as (3S, 4S, 6S). Thus the two stereoisomers differed in their configuration at C-3 and C-4 rather than at C-6, a result that has considerable biosynthetic importance (vide infra). Plakortolides K and L, both with a C12 methylene chain, represent homologues of previously published plakortolides that contain a C10 methylene chain. The absolute configuration of plakortolide E (17), isolated from a Fijian specimen of Plakortis, was deduced as (3R, 4R, 6R) by Crews et al. using Mosher ester analysis.6 A plakortolide metabolite (18) from a Philippine sponge Plakinastrella sp., with structure and absolute configuration assigned as (3S, 4S, 6R) by Faulkner et al.,7 was recently subjected to total synthesis by Jung et al. using a biomimetic strategy.12 These authors named the metabolite as 6-epiplakortolide E even though it possesses the same 6R configuration as plakortolide E.12 Plakortolide K (1) has an [R]D of þ8.8, equal and opposite to the literature [R]D value of -8 reported for 18 by Faulkner et al.,7 while the NMR data for the core regions matched in all respects. The [R]D and Mosher results confirm that these two plakortolides have the same relative configuration but are opposite in absolute configuration. In contrast, when we compared the data for plakortolide L (2) to that of the supposedly homologous structure 17 named plakortolide E,6

differences in the NMR data for the bicyclic core were immediately apparent, in particular for C-3, C-6, and the C-6 Me group (2: C-3, δH 4.43, d; δC 81.0; C-6 δC 79.9; and C-24 δH 1.27; δC 22.2; for “plakortolide E”: C-3: δH 4.19, d; δC 73.7; C-6 δC 72.9; and (using our numbering scheme) C-24 δH 1.35; δC 29.9). The published NMR data provided by Crews et al.6 for the Fijian “plakortolide E” metabolite did not match the actual structure 17 that they had reported. Instead, following comparison with NMR data for synthetic diols 10 and 11, it was apparent that the NMR data were in fact for diol 19, equivalent to 11 in all respects except for alkyl chain length and absolute configuration. Serendipitously, a polar fraction from the P. clathrata extract showed evidence of 1H NMR signals that were reminiscent of those reported for the Fijian “plakortolide E” sample.6 Following RP-HPLC purification, the 1H NMR and HSQC data (Tables 3 and 4) for the isolated metabolite showed a phenyl ring, one oxygenated methine proton (δH 4.17; δC 73.8), two distinctive AB systems (δH 2.90, 2.52; δC 38.1 and δH 2.16, 2.07; δC 43.9), two methyl singlets (δH 1.42, 1.34; δC 26.9, 30.2), and a cluster of overlapping methylene signals for a C12 alkyl chain. The HRESIMS data revealed a formula of C26H42O4 that supported a reduced plakortolide structure, and when the HMBC evidence revealed a lactone (δC 175.0), the seco-plakortolide L (diol) structure 11 was deduced. The 1H NMR data for the spongederived seco-plakortolide L were identical to those obtained for the sample prepared by Zn/AcOH reduction of 2. Analogously, the sponge extract also yielded a small sample of seco-plakortolide K (10), whose 1H NMR data again matched those of the synthetic sample prepared from plakortolide K (1). Owing to the small quantities available following purification by RP-HPLC, optical rotation data were not obtained for the sponge-derived samples of diols 10 and 11. Indeed, during measurements on the samples of seco-plakortolides 10 and 11 synthesized from 1 and 2, we had noted that these two compounds did not give reliable optical rotation readings. The difficulty in obtaining data was not related to any chemical instability of the diols since recovery of samples after [R]D measurements led only to seco-plakortolide products. This is despite the seemingly facile conversion of secoplakortolides to plakortone ethers, as is explained further below. 198

dx.doi.org/10.1021/np100620x |J. Nat. Prod. 2011, 74, 194–207

Journal of Natural Products

ARTICLE

Chart 2

Table 3. 1H NMR Assignments for Sponge-Derived seco-Plakortolides 10-13 and Synthetic Diols 22 and 23a 10b

11b

12b

13b

22c

23c

2a

2.88, dd (18.2, 6.2)

2.90, dd (18.0, 6.5)

2.88, dd (18.0, 6.4)

2.90, dd (18.0, 6.4)

2.89, dd (18.5, 6.0)

2.91, dd (18.5, 6.5)

2b

2.50, dd (18.2, 1.0)

2.52, dd (18.0, 2.0)

2.50, d (18.0)

2.52, d (18.0)

2.51, d (18.5)

2.53, dd (18.5, 1.5)

3

4.21, dd (6.2, 1.0)

4.17, dd (6.5, 2.0)

4.21, d (6.4)

4.17, d (6.4)

4.22, d (6.0)

4.17, dd (6.5, 1.5)

5a

2.36, d (14.9)

2.16, d (15.0)

2.36, d (15.0)

2.15, d (15.0)

2.37, d (14.5)

2.15, d (15.0)

position

5b

1.86, d (14.9)

2.07, d (15.0)

1.86, d (15.0)

2.08, d (15.0)

1.86, d (14.5)

2.09, d (15.0)

7 8

1.57-1.51, m 1.28, m

1.62-1.51, m 1.28, m

1.55-1.50, m 1.28, m

1.55-1.50, m 1.28, m

1.54, m 1.28, m

1.52-1.61, m 1.28, m

9

1.23-1.29, md,e

1.24-1.29, md,e

1.23-1.30, md,e

1.23-1.30, md,e

1.24-1.28, md,e

1.24-1.28, md,e

d,e

d,e

d,e

d,e

d,e

1.24-1.28, md,e

d,e

1.24-1.28, md,e

d,e

1.24-1.28, md,e

d,e

1.24-1.28, md,e

d,e

1.24-1.28, md,e

d,e

1.24-1.28, md,e 1.24-1.28, md,e

10 11 12 13 14

1.23-1.29, m

d,e

1.23-1.29, m

d,e

1.23-1.29, m

d,e

1.23-1.29, m

d,e

1.23-1.29, m

d,e

1.24-1.29, m

d,e

1.24-1.29, m

d,e

1.24-1.29, m

d,e

1.24-1.29, m

d,e

1.24-1.29, m

d,e

1.23-1.30, m

1.23-1.30, m

d,e

d,e

1.23-1.29, m

1.23-1.30, m

d,e

d,e,e

1.23-1.30, m

1.23-1.30, m

d,e

d,e

1.23-1.30, m

1.23-1.30, m

2.12, q (7.0)

2.12, q (7.0)

5.81, dt (15.0, 7.0) 6.18, dd (15.0, 10.0)

5.81, dt (15.0, 7.0) 6.18, dd (15.0, 10.0)

1.24-1.28, m 1.24-1.28, m 1.24-1.28, m 1.24-1.28, m

1.24-1.28, m

15 16

1.23-1.29, m 1.23-1.29, md,e

1.24-1.29, m 1.24-1.29, md,e

1.24-1.28, m 1.24-1.28, md,e

17

1.59, m

1.58, m

6.73, dd (15.8, 10.0)

6.73, dd (15.8, 10.0)

1.55, m

1.55, m

18

2.58, t (7.8)

2.58, t (7.8)

6.42, d (15.8)

6.42, d (15.8)

2.51, t (7.8)

2.51, t (7.8)

20

7.16, m

7.16, m

7.35, d (7.5)

7.35, d (7.5)

7.01, d (8.5)

7.01, d (8.5)

21

7.25, m

7.25, m

7.27, t (8.0)

7.27, t (8.0)

6.72, d (8.5)

6.72, d (8.5)

22

7.15, m

7.15, m

7.17, t (7.5)

7.17, t (7.5)

23

1.45, s

1.42, s

1.45, s

1.42, s

1.46, s

1.42, s

24 22-OH

1.34, s

1.33, s

1.34, s

1.34, s

1.35, s 4.80, br s

1.34, s 4.80, br s

a Chemical shifts (ppm) referenced to CHCl3 (δC 7.24). b At 750 MHz. c At 500 MHz. d Unresolved chemical shifts due to overlapping signals. e Signal multiplicity unresolved due to overlapping signals.

199

dx.doi.org/10.1021/np100620x |J. Nat. Prod. 2011, 74, 194–207

Journal of Natural Products Table 4. position

13

ARTICLE

C NMR Assignments for Sponge-Derived seco-Plakortolides 10-13 and Synthetic Diols 22 and 23a 10b

11b

12b

13b

22c

23c

1

174.6

175.0

174.6

174.7

174.9

175.1

2

37.6

38.1

37.6

37.9

37.5

37.9

3

73.3

73.8

73.4

73.8

73.3

73.7

4

90.0

89.7

90.1

89.7

90.3

89.8

5

43.9

43.9

43.9

43.7

43.8

43.6

6

72.5

73.1

72.6

73.1

72.6

73.1

7

46.4

43.7

46.5

43.6

46.4

43.5

8 9

22.5 29.0-29.9e

24.3 29.0-29.9e

-d 29.0-29.9e

-d 29.0-29.9e

22.8 29.1-29.2e

24.1 29.3-29.6e

10

29.0-29.9e

29.0-29.9e

29.0-29.9e

29.0-29.9e

29.1-29.2e

29.3-29.6e

11

e

e

e

e

e

29.3-29.6e

e

29.3-29.6e

e

12

29.0-29.9

e

29.0-29.9

e

29.0-29.9

e

29.0-29.9

e

29.0-29.9

e

29.0-29.9

e

29.0-29.9

e

29.0-29.9

e

29.1-29.2 29.1-29.2

13

29.0-29.9

29.0-29.9

29.0-29.9

29.0-29.9

29.1-29.2

29.3-29.6e

14

29.0-29.9e

29.0-29.9e

32.7

32.7

29.1-29.2e

29.3-29.6e

15

29.0-29.9e

29.0-29.ee

135.7

135.7

29.1-29.2e

29.3-29.6e

16 17

29.0-29.9e 31.3

29.0-29.9e 31.4

130.4 129.3

130.4 129.3

29.1-29.2e 31.3

29.3-29.6e 31.4

18

35.8

36.1

129.8

129.8

34.7

34.7

19

142.6

142.9

137.4

137.4

134.9

135.0

20

128.2

128.3

125.9

125.9

129.2

129.3

21

128.0

128.2

128.3

128.3

114.8

114.9

22

125.4

125.5

126.9

126.9

153.4

153.4

23

25.9

26.9

26.0

26.7

25.9

26.7

24

28.2

30.2

28.3

30.0

28.2

29.8

a

Chemical shifts (ppm) taken from 2D spectra referenced to CDCl3 (δC 77.0). b At 750 MHz. c At 500 MHz. d Not detected. e Unresolved chemical shifts taken from HSQC experiments.

measured as -11 (c 0.08, CHCl3) and -13 (c 0.05, CHCl3), respectively, illustrating the experimental difficulties associated with [R]D measurements for this group of compounds. Four metabolites isolated from the Plakinastrella extract with a plakortolide-like core showed evidence of diene functionality by UV data; they were carefully purified by repetitive RP-HPLC. Two diastereomers, 5 and 6, each with a C12 diene side chain from MS data and corresponding to a molecular formula of C26H36O4, were named plakortolides O and P, respectively, while metabolites 7 and 8, named plakortolides Q and R, both had a C10 diene side chain from MS data, corresponding to a molecular formula of C24H32O4. 1H NMR signals at δH 5.81 (dt, H-15), 6.18 (dd, H-16), 6.73 (dd, H-17), and 6.42 (dt, H-18) were observed for the diene side chain of 5, and equivalent signals were observed in the 1H NMR spectra of 6-8. There was insufficient 5 for 13C NMR characterization, but for 6 HMBC signals from the olefinic protons H-17 and H-18 to aromatic carbons at δC 137.4 (C-19) and δC 125.9 (C-20), from the H-20 signal at δH 7.35 to the olefinic carbon at δC 129.8 (C-18), from H-18 to δC 130.4 (C-16), and from H-17 to δC 135.8 (C-15) all secured the conjugated diene next to the aromatic ring. Equivalent HMBC correlations were seen for 7 and 8. Isomers 5 and 7 showed 1H NMR and 1H/13C NMR data, respectively, for the bicyclic core similar to 1 and 3, while the data for 6 and 8 matched those of 2 and 4. In 7, there were NOE correlations from H3-24 to H-5b and from H-5a to H2-7, as had been found earlier for 1. In contrast, 6 and 8 both showed equivalent NOE correlations to 2, including from H3-24 to H-5a. For each of 5-8, the conjugated diene was assigned a (E,E)-configuration from coupling constant data. The samples of 5-8 were individually